Background In 1999 our laboratory discovered that mutations in TNFRSF1A, the gene encoding the 55 kDa tumor necrosis factor (TNF) receptor, cause a dominantly inherited syndrome of periodic fever and inflammation. Although one of the first families reported with this condition was of Irish ancestry, and had therefore been denoted "familial Hibernian fever," we found TNFRSF1A mutations in patients of several different backgrounds, and we therefore proposed the more ethnically neutral name "TNF receptor associated periodic syndrome." For the C52F mutation, we demonstrated a defect in activation-induced ectodomain shedding of the TNF receptor, which may interfere with normal homeostasis in inflammation. During the 3 years leading to the present reporting period, we (1) performed additional mutational studies, bringing the total number of known mutations to 17 (13 of which were first identified in our laboratory); (2) demonstrated increased risk for amyloidosis among TRAPS patients with mutations at cysteine residues; (3) demonstrated that the R92Q variant is present in approximately 1% of Caucasian control chromosomes, and is seen at increased frequency in an early arthritis clinic; (4) confirmed p55 receptor shedding defects for most, but not all, mutations tested, suggesting that this is not the only mechanism of disease; (5) initiated a treatment protocol to evaluate the efficacy of the TNF-blocking agent etanercept in TRAPS; and (7) began developing TRAPS knockin mice. Late in 2001, another laboratory reported mutations in CIAS1, which encodes a homolog of the protein that causes familial Mediterranean fever, in two dominantly inherited periodic fevers, Muckle-Wells syndrome (MWS) and familial cold autoinflammatory syndrome (FCAS). Results of the Last Year Mutational and clinical observations: During the last year we have identified two novel mutations in TNFRSF1A, C30Y and S74C. We have also performed detailed imaging and histologic studies on a patient with TRAPS migratory myalgia, and have found only fasciitis, but no evidence of myositis. Studies of the molecular pathogenesis of TRAPS: During the last reporting period, we developed expression vectors for wild-type TNFRSF1A, and for H22Y, C30S, C33G, T50M, C52F, C88R, and R92Q mutations. In a 293T transfection system, we found that biotinylated TNF binds R92Q mutants as well as the wild-type receptors, that TNF binding to H22Y and C30S is diminished but not abolished, and TNF binding to the other 4 mutants is totally ablated. We have now found that cotransfection of the wild-type receptor to simulate the situation in heterozygous TRAPS patients results in complete normalization of TNF binding. Moreover, using both fluorescence resonance energy transfer (FRET) and a dominant-negative apoptosis inhibition assay, we found that at least the 4 cysteine mutants do not interact with the wild-type receptor. Development of TRAPS knockin animal models: To date we have successfully bred heterozygous and homozygous mice with the T50M mutation, and have documented heterozygous germline transmission of the C33Y mutation. The C52F construct is awaiting blastocyst injection. Neither the homozygous nor the heterozygous T50M knockin mice have any developmental abnormalities, but more sensitive studies of their phenotype are now under way. Studies of TNF-inhibition in the treatment of TRAPS: We have now completed an open-label study of the TNFR p75:Fc fusion protein, etanercept, in TRAPS. The trial consisted of a 3-month observation period on "standard therapy" (usually intermittent corticosteroids), 3 months on etanercept twice weekly (25 mg sc for adults, 0.4 mg/kg for children), 3 months on etanercept 3 times a week if there had not been a complete response to twice weekly (otherwise, a continuation of etanercept twice weekly), and a 3 month washout period. Patients were monitored for the number and severity of attacks, baseline and attack-associated levels of a number of inflammatory parameters, and the need for other medications (analgesics, corticosteroids, etc.). Of the 15 patients enrolled, 10 had the T50M mutation, 2 had H22Y, 1 had P46L, 1 had R92Q, and 1 had C33G. None had evidence of systemic amyloidosis. There were 2 withdrawals: the R92Q patient withdrew for lack of efficacy, and the C33G patient withdrew in the first quarter because of inability to undergo frequent blood testing. Of the remaining 13, all required a dose escalation in the third quarter. There were no serious adverse events. Subjects experienced a sharp improvement in an "attack score" from the first to the second quarter (p<0.0001), a smaller improvement from the second to the third quarter (p<0.0015), and a sharp rebound in the fourth quarter (p<0.0001). Moreover, subjects had a dose dependent decrease in their use of other medications, and experienced a significant reduction in the erythrocyte sedimentation rate, C-reactive protein, and serum amyloid A levels while on etanercept. Discovery of de novo dominant mutations in CIAS1 in patients with NOMID/CINCA: Based on the similarity between a patient with neonatal onset multisystem inflammatory disease (NOMID; also known as chronic infantile neurologic, cutaneous, and articular syndrome, or CINCA) and MWS, we sequenced CIAS1 in the NOMID/CINCA patient. We found a G -> A transition resulting in the substitution of asparagine for aspartic acid at codon 303 (D303N) that was not present in any of 1126 ethnically-matched control chromosomes. Additional samples were obtained, and overall in 6 of 13 patients we found heterozygous missense substitutions in CIAS1. Five of the mutations (L264H, A374N, Y570C, and two distinct nucleotide substitutions causing F523L) were novel. None of the sequence changes was observed in a panel of over 900 chromosomes from healthy controls. In 4 mutation positive children from whom parental DNA was available, no mutation was found in the parental DNA, supporting the hypothesis that the mutations arose de novo. Consistent with the recently discovered role of CIAS1 in IL-1 regulation, we found evidence for increased IL-1 beta, as well as TNF, IL-3, IL-5, and IL-6, but not TGF beta, in a mutation-positive patient compared with normal controls. Conclusions and Significance Our data continue to expand the mutational spectrum of TRAPS. Our studies of transfected mutant TNF receptors substantially change our concept of how TNFRSF1A mutations in TRAPS work. Our current concept is that cysteine-mutant receptors may be sufficiently structurally altered that they don't bind ligand on the cell surface, don't compete for TNF in solution, can't interact with wild-type receptors, and aren't recognized by the enzymes that would cleave them from the cell surface. This may lead to excessive inflammation by failing to transmit a negative signal through the p55 receptor, or shunting down a potentially stimulatory p75-dependent pathway. Our etanercept data support the clinical use of this agent in TRAPS, and indicate that TRAPS inflammation is ligand-dependent. Our data also firmly establish mutations in CIAS1 as a cause of NOMID/CINCA. Besides the obvious diagnostic implications, our data place this disorder at the severe end of a spectrum of CIAS1-associated inflammatory disorders, and suggest that therapies targeted at IL-1 may be beneficial for these patients. During the next year, our objectives are: (1) to continue mutational and genotype-phenotype studies of TRAPS; (2) to continue studies of cell lines transfected with TNFRSF1A mutants; (3) to begin physiologic studies of TRAPS knockin mice; (4) to search for mutations in genes in the CIAS1 family and pathway in mutation-negative patients with NOMID/CINCA; and (5) to undertake therapeutic studies of anakinra, the IL-1 receptor antagonist,